The inputs for developmental transcription networks are the differential concentrations of transcription factor proteins in each nucleus of the embryo over time. The outputs are the levels of transcription of target genes in each nucleus. Simple visual inspection of embryos stained to reveal gene expression patterns shows that many genes' expression levels differ from cell to cell in a highly complex spatial and temporal manner (e.g. Fig. 4 in the Program Overview [1]). Therefore, to be able to model how combinations of transcription factors coordinate to produce such complex output patterns, it is essential to quantitate factor protein and target mRNA expression in 3D with cellular resolution. To this end, we have developed image analysis based methods, described below, that provided the first description of morphology and quantitative gene expression of a complete embryo with cellular resolution in a computationally analyzable format, a VirtualEmbryo [2-5]. Project 2 will generate hundreds of transgenic fly lines in which CRMs bearing point mutations in transcription factor recognition sites are placed upstream of a proximal promoter/reporter gene construct. It is likely (and highly desirable for testing our models) that most of these mutations will lead to modest (several fold) changes in occupancy of 1-3 transcription factors and concomitant quantitative changes in parts of the expression patterns driven by the CRMs. To be able to measure these transcription changes, it will be necessary to use our image analysis pipeline to quantitate the spatial and temporal patterns of reporter gene expression driven by both the original and the mutated forms of these CRMs. We will do this in this Aim.